Variable importance matching for causal inference

Quinn Lanners, Harsh Parikh, Alexander Volfovsky, Cynthia Rudin, David Page
Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence, PMLR 216:1174-1184, 2023.

Abstract

Our goal is to produce methods for observational causal inference that are auditable, easy to troubleshoot, yield accurate treatment effect estimates, and scalable to high-dimensional data. We describe a general framework called Model-to-Match that achieves these goals by (i) learning a distance metric via outcome modeling, (ii) creating matched groups using the distance metric, and (iii) using the matched groups to estimate treatment effects. Model-to-Match uses variable importance measurements to construct a distance metric, making it a flexible framework that can be adapted to various applications. Concentrating on the scalability of the problem in the number of potential confounders, we operationalize the Model-to-Match framework with LASSO. We derive performance guarantees for settings where LASSO outcome modeling consistently identifies all confounders (importantly without requiring the linear model to be correctly specified). We also provide experimental results demonstrating the auditability of matches, as well as extensions to more general nonparametric outcome modeling.

Cite this Paper

BibTeX
@InProceedings{pmlr-v216-lanners23a, title = {Variable importance matching for causal inference}, author = {Lanners, Quinn and Parikh, Harsh and Volfovsky, Alexander and Rudin, Cynthia and Page, David}, booktitle = {Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence}, pages = {1174--1184}, year = {2023}, editor = {Evans, Robin J. and Shpitser, Ilya}, volume = {216}, series = {Proceedings of Machine Learning Research}, month = {31 Jul--04 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v216/lanners23a/lanners23a.pdf}, url = {https://proceedings.mlr.press/v216/lanners23a.html}, abstract = {Our goal is to produce methods for observational causal inference that are auditable, easy to troubleshoot, yield accurate treatment effect estimates, and scalable to high-dimensional data. We describe a general framework called Model-to-Match that achieves these goals by (i) learning a distance metric via outcome modeling, (ii) creating matched groups using the distance metric, and (iii) using the matched groups to estimate treatment effects. Model-to-Match uses variable importance measurements to construct a distance metric, making it a flexible framework that can be adapted to various applications. Concentrating on the scalability of the problem in the number of potential confounders, we operationalize the Model-to-Match framework with LASSO. We derive performance guarantees for settings where LASSO outcome modeling consistently identifies all confounders (importantly without requiring the linear model to be correctly specified). We also provide experimental results demonstrating the auditability of matches, as well as extensions to more general nonparametric outcome modeling.} }
Endnote
%0 Conference Paper %T Variable importance matching for causal inference %A Quinn Lanners %A Harsh Parikh %A Alexander Volfovsky %A Cynthia Rudin %A David Page %B Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2023 %E Robin J. Evans %E Ilya Shpitser %F pmlr-v216-lanners23a %I PMLR %P 1174--1184 %U https://proceedings.mlr.press/v216/lanners23a.html %V 216 %X Our goal is to produce methods for observational causal inference that are auditable, easy to troubleshoot, yield accurate treatment effect estimates, and scalable to high-dimensional data. We describe a general framework called Model-to-Match that achieves these goals by (i) learning a distance metric via outcome modeling, (ii) creating matched groups using the distance metric, and (iii) using the matched groups to estimate treatment effects. Model-to-Match uses variable importance measurements to construct a distance metric, making it a flexible framework that can be adapted to various applications. Concentrating on the scalability of the problem in the number of potential confounders, we operationalize the Model-to-Match framework with LASSO. We derive performance guarantees for settings where LASSO outcome modeling consistently identifies all confounders (importantly without requiring the linear model to be correctly specified). We also provide experimental results demonstrating the auditability of matches, as well as extensions to more general nonparametric outcome modeling.
APA
Lanners, Q., Parikh, H., Volfovsky, A., Rudin, C. & Page, D.. (2023). Variable importance matching for causal inference. Proceedings of the Thirty-Ninth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 216:1174-1184 Available from https://proceedings.mlr.press/v216/lanners23a.html.

Related Material